I’ve already installed the cameras, only to realize I didn’t understand the model number?

“Brother, this...”MV–CL084–90CMWhat does it mean, exactly?

What’s most frightening about being on-site isn’t that the problems are difficult. It’s that you think you understand, but as soon as you start working on the actual system, you run into one pit after another.

In industrial vision systems,Line-scan cameraHe’s definitely a tough guy.

High speed. High resolution. Particularly well-suited for targets in high-speed, continuous motion.

Therefore, it is commonly used in applications such as orbital inspection, lithium battery testing, printing, textiles, and semiconductors.

But problems have also arisen.

A line-scan camera isn't something you can just “buy a high-end model and call it good.” Every character in the model number could determine whether your subsequent debugging goes smoothly or requires rework on-site.

1. Don't underestimate the model—there's a selection code hidden inside.

Let’s take this model as an example:

At first glance, it looks like a string of product codes. In reality, when you take it apart, each segment has a clear meaning.

Do you understand?

This isn’t just a random name chosen by the manufacturer. It’s a model number that packs in the product type, camera structure, resolution, sensor, interface, and image format—all into one single designation.

When selecting a model on-site, if you can’t understand the model specifications, it’s like holding a map but not knowing where you are. It may seem like you can still make progress, but in reality, it’s easy to veer off course.

2. MV, C, L: First, figure out exactly what kind of camera it is.

First, look at the first few characters.

MVrepresenting machine vision products. These products are primarily geared toward applications such as industrial automation and quality inspection.

CIt refers to an industrial camera. In other words, it’s not a typical consumer-grade camera but rather a high-performance product designed for industrial applications.

Next comes the key.L.

Here is...L, representsLine Scan Camera.

It's different from common area-array cameras.

A two-dimensional (2D) camera captures an entire image in a single shot. A line-scan camera, on the other hand, typically features a single-row or multi-row pixel structure and is well-suited for capturing continuously moving objects.

This explains why it is commonly used in high-precision inspection scenarios such as steel material defects, PCB flaws, and print registration.

The advantages of line-scan cameras are straightforward:

It can run at high speeds. It can achieve high resolution. It’s suitable for continuous imaging.

But it’s also very sensitive to site conditions. Not a single factor—such as wire speed, detection width, resolution requirements, or lighting setup—can be decided arbitrarily on the spur of the moment.

At this point, saying “Just adjust the parameters” is basically like digging your own grave.

3. 084 Not a number, but a resolution clue.

Many engineers are seeing it for the first time.084It’s easy to mistake it for a regular serial number.

No.

It represents the camera’s effective imaging resolution, which is:

Number of pixels × Number of channels

Common encodings can be understood as follows:

So, in the example:084Indicates:

8192Pixels × 4 rows, or 4 channels.

This type of configuration is suitable for high-resolution, high-speed detection tasks—particularly in scenarios that demand highly precise capture of target details.

There’s often a saying at the scene:

“How come the defect slipped through again?”

Don't rush to blame the algorithm—or to suspect that your camera is broken.

You first need to ask:Is the resolution sufficient? Have you captured enough imaging detail?

After tweaking the parameters for ages, it turned out the root cause wasn't in the algorithm. In visual projects, this kind of thing isn't unusual at all.

4. Sensors, interfaces, black-and-white versus color—these are the real keys to successful implementation.

Keep looking ahead.

In the model4, indicating the sensor manufacturer's code.

In this model,4 Indicates adoption.CMOSIS linear array image sensor.

The sensor is the core component of an industrial camera. It affects key performance metrics such as image quality, sensitivity, and noise control.

Next is 9.

This location is a reserved slot. Currently, no specific function has been assigned to it; it is primarily intended for future expansion models or compatibility placeholders.

When selecting a model, you can ignore this one. Don't get hung up on minor details here.

What really needs to be closely monitored is the interface behind it.

In the modelC, representsCamera Link Interface.

The characteristics of Camera Link are:

High-speed transmission. Stable and reliable. Suitable for high-frequency acquisition scenarios in industrial cameras.

But please note that an interface isn't just about looking good on paper. It must be compatible with the capture card or system motherboard.

If the interfaces don’t match, it’s not “tuning parameters” on-site—it’s “starting from scratch.”

The last oneM, representing black-and-white images.

The example inM, indicating that this is a...Black-and-white line-scan cameraIt is suitable for detection tasks with distinct texture contrasts.

Whether to use black-and-white or color doesn't depend on whether the customer likes it or not. Instead, it depends on the detection target, defect characteristics, and imaging requirements.

What visual projects fear most isn't difficulty—it's getting the direction wrong from the very beginning.

5. Back to the scene: How exactly should we interpret MV–CL084–90CM?

Now let’s take another look at this model:

MV–CL084–90CM

It can be broken down like this:

In other words, this isn't a camera you'd just “browse through” casually.

It is an industrial line-scan camera belonging to Hikvision’s machine vision product lineup, with a resolution code of 084, corresponding to a line-scan array of 8192 × 4 pixels. It features a CMOSIS sensor and uses the Camera Link interface to output black-and-white images.

Once you understand this series of model numbers, selecting the right one won't rely solely on intuition.

You can at least quickly determine:

Is it a line array? Is its resolution level sufficient? Can it be connected to the interface? Is it black and white or color? Does it meet the requirements for high-speed inspection tasks?

If these questions aren't clarified first, they'll turn into a series of follow-up inquiries on the spot later on.

“Why can’t the image be transmitted?” “Why can’t the details be captured clearly?” “Why is the capture card incompatible?” “Why does the customer still say defects are being missed?”

The camera isn’t broken. The light source isn’t broken. What’s broken is our initial judgment.

When selecting a line-scan camera, don't just focus on whether it's "expensive or not."

When selecting a line-scan camera, don't just focus on the brand, price, or resolution. And certainly don't place an order simply because you see the words "high performance."

The truly reliable approach is to make judgments based on actual application scenarios:

What is the linear velocity? What’s the detection width? How high is the required resolution? Can the lighting solution be properly matched? Are the interface and acquisition system compatible? Should we go with black-and-white or color? Between GigE, USB3 Vision, and Camera Link, which one is more suitable?

The model naming convention isn't esoteric knowledge—it’s the first hurdle to avoid pitfalls on the engineering site.

Only by understanding the model specifications can you quickly narrow down the options to those that meet your bandwidth, image quality, and interface protocol requirements. When you combine this with the project’s specific needs, your selection process will become much more efficient.